Peritoneal dialysis adequacy: A model to assess feasibility with various modalities

Transcription

1 Kidney International, Vol. 55 (1999), pp Peritoneal dialysis adequacy: A model to assess feasibility with various modalities JOSE A. DIAZ-BUXO, FRANK A. GOTCH, TOM I. FOLDEN, SHELDEN ROSENBLUM, JAMES ZAZRA, NANCY LEW, TERRI L. CRAWFORD, BRENDA YOUNGBLOOD, ANGIE PESICH, and J. MICHAEL LAZARUS Fresenius Medical Care North America, Lexington, Massachusetts; Dialysis Treatment and Research, Davies Medical Center, San Francisco, and Fresenius USA, Walnut Creek, California; and LifeChem Laboratories, Rockleigh, New Jersey, USA Peritoneal dialysis adequacy: A model to assess feasibility with weekly normalized urea clearance (Kprt/V) of 2.0 or various modalities. more and a creatinine clearance (C Cr ) of 60 liter/1.73 m 2 Background. The current standard of adequacy for peritoor more [3]. Recognized weaknesses of these guidelines neal dialysis (PD) is to provide a weekly normalized urea clearance (Kt/V) of 2.0 or more and a creatinine clearance include: (a) the lack of evidence to establish equivalency (C Cr ) of 60 liter/1.73 m 2 or more. As native renal function is lost, between the renal and peritoneal contributions of small it is important to determine the effectiveness of the available solute removal to clinical outcome; and (b) the uncertain therapeutic modalities in achieving these goals. value of urea versus creatinine as markers of uremia Methods. A model to assess our ability to provide a weekly [4, 5]. Most clinical outcome studies have included pa- Kt/V urea of 2.0 or more and a C Cr of 60 liter/1.73 m 2 or more to anuric patients undergoing continuous ambulatory PD tients at various stages of uremia therapy with different (CAPD) and automated PD (PD Plus) was developed. The degrees of residual renal function (RRF). The most body surface area (BSA) distribution was obtained from 38,768 quoted prospective study, CANUSA, was not designed patients undergoing dialysis during January The distribu- to evaluate the relative effect of RRF and the peritoneal tion of peritoneal transport rates (PTRs) was obtained from clearance on patient outcome [2, 6, 7]. Furthermore, 2531 peritoneal equilibration tests performed during The weekly K p t/v urea was calculated for the various PTR groups patients with significant RRF are more likely to achieve and the range of BSA with four PD prescriptions: CAPD 8 higher clearances of creatinine than anuric patients be- liters, CAPD 10 liters, PD Plus 12 liters, and PD Plus 15 liters, cause of tubular secretion of creatinine in advanced renal using a previously validated kinetic program (PackPD). failure. Few prospective studies have been performed in Results. The predicted percentage of patients capable of achieving the adequacy goals for Kt/V and C anuric patients undergoing PD. However, Selgas et al Cr, respectively, were 24.8 and 11.2 for CAPD 8 liters, 54.2 and 33.0 for CAPD studied patients who had a minimum of three years on 10 liters, 77.8 and 54.9 for PD Plus 12 liters, and 93.2 and 72.9 PD and who were mostly anuric [4]. They observed an for PD Plus 15 liters. excellent correlation between Kt/V urea and survival, but Conclusions. Most patients can attain the current adequacy failed to establish a correlation between C Cr and outcome. standards of therapy with automated PD, but few (less than 25%) can do so with standard CAPD in the absence of residual As RRF is lost, the dose of PD must be necessarily renal function. adjusted in most patients undergoing continuous ambu- latory PD (CAPD) and many undergoing automated PD (APD) in order to satisfy the adequacy criteria. The New standards for peritoneal dialysis (PD) adequacy proportion of patients capable of achieving the aforehave been recently proposed based on theoretical conhas not been established. Theoretical models have con- mentioned goals of PD adequacy in the absence of RRF structs and clinical outcome studies [1 3]. These recomsidered the patients peritoneal transport rates (PTRs), mendations are to provide a total (peritoneal and renal) body size, and dialysis prescription and have either disregarded ultrafiltration or modeled it from their peritoneal Key words: kinetic modeling, continuous ambulatory peritoneal dialequilibration tests (PETs) [8 10]. However, they lacked ysis, automated peritoneal dialysis, urea clearance, creatinine clearance, residual renal function. statistical information on the distribution of patient size, distribution of PTRs, or actual solute removal with spe- Received for publication September 9, 1998 and in revised form January 11, 1999 cific PD modalities. Accepted for publication January 12, 1999 This study assesses the feasibility of achieving PD adequacy with various therapeutic modalities by using 1999 by the International Society of Nephrology statis- 2493

2 2494 Diaz-Buxo et al: Peritoneal dialysis adequacy METHODS In order to develop a model to assess the proportion of anuric patients capable of achieving a weekly Kpt/V of 2.0 or more or a C Cr of 60 liter/1.73 m 2 or more, the following information is required concerning the population to be studied: (a) the distribution of patient size, (b) the distribution of PTRs, and (c) the clearances obtained with various prescriptions in patients with different body size and PTRs. Determination of patient size distribution The records of all patients undergoing hemodialysis (HD) three times weekly on the consecutive Monday and Tuesday of the third week of January 1997, and of all PD patients seen for their monthly visit during the same month in all Fresenius Medical Care North America facilities were used to determine gender, height (cm), and weight (kg). For HD patients, the postdialysis weight was used. For PD patients, the weight used was the only one recorded for that visit; however, there was not sufficient information to determine whether the patients were weighed during dialysis or after a full drain. Body surface area (BSA) was calculated using du Bois and du Bois formula in 35,103 HD and 3665 PD patients with complete records [11]. Distribution histograms were generated for each population from these data. Determination of peritoneal transport rate distribution The distribution of PTR was based on data from PETs performed at LifeChem Laboratories (Rockleigh, NJ, USA) between January 1, 1996, and December 31, 1996, on 2531 patients. Standard methodology was used for four-hour PET determinations [12]. Only complete studies with data at times 0, 2, and 4 hours were included. Dialysate creatinine values were corrected for glucose Fig. 1. Distribution of body surface area (BSA) for (A) hemodialysis, concentration [12, 13]. Plasma concentrations for urea (B) peritoneal dialysis, and (C) a combined group of dialysis patients. and creatinine were used. The PTR classification was based on four-hour dialysate to plasma ratios (D/P 4hr ) for creatinine as described by Twardowski et al [12]. Four groups of PTR were defined as follows: low, for Table 1. Classification of peritoneal transport groups values between 1% and mean 1 sd; low-average, be- Group D/P Cr4hr D 4 D 0 tween mean and mean 1 sd; high-average, between Low mean and mean 1 sd; and high, between mean 1 sd Low-average and 99%. High-average High tical information for patient size and peritoneal transport generated during a recent period of time with a large number of patients who were considered representative of the United States PD end-stage renal disease (ESRD) population and kinetic modeling using a validated program. Definition of peritoneal dialysis modalities and prescriptions Four PD prescriptions were used for modeling using various total dialysis solution volumes. Two prescriptions used manual CAPD and two used a hybrid APD-CAPD modality known as PD Plus [14]. PD Plus consists of three to five automated cycles at night that are delivered by a simple peritoneal cycler and use high exchange volumes (Vip) in the range of 2.5 to 4.0 liters. The first

3 Diaz-Buxo et al: Peritoneal dialysis adequacy 2495 Table 2. Definition of peritoneal transport curves Peritoneal transport group Prescription Low Low-average High-average High CAPD-8 y x y x y x y x CAPD-10 y x y x y x y x PD 12 y x y x y x y x PD 15 y x y x y x y x Fig. 2. Relationship between weekly Kpt/V urea and volume of distribution of urea (Vsa) according to four different peritoneal dialysis prescriptions and peritoneal transport rates. Symbols are: ( ) PD Plus-15; ( ) PD Plus-12; ( ) CAPD-10; ( ) CAPD-8. diurnal exchange is the last exchange delivered by the The four prescriptions used were as follows: CAPD cycler in the morning. An additional manual exchange with four 2-liter exchanges (CAPD-8); CAPD with four is provided at midday to avoid prolonged dwell times in 2.5-liter exchanges (CAPD-10); PD Plus using three excess of seven hours. The diurnal Vip varies between 3-liter exchanges at night and two, 1.5-liter exchanges 1.5 and 2.0 liters for most adult patients. during the day (PD Plus-12); and PD Plus using four

4 2496 Diaz-Buxo et al: Peritoneal dialysis adequacy insufficient values. Data from actual C Cr determinations from 188 patients representing the four PTR groups and who were undergoing PD with one of the four different prescriptions were used instead. The C Cr values reported were normalized to BSA (liter/1.73 m 2 ). Fig. 3. Relationship between weekly C Cr and peritoneal transport rates for four different prescriptions. The figures in parenthesis denote the number of patients studied. A description of the dialysis prescriptions is in the text. Symbols are: ( ) CAPD-8; ( ) CAPD-10; ( ) PD Plus-12; ( ) PD Plus liter exchanges at night and two 2.0-liter exchanges during the day (PD Plus-15). All prescriptions were modeled using dialysis solutions with 2.5% glucose. Determination of clearance with different prescriptions The frequency distribution of the four PTR groups (fptr) in the population is known. We also know the cumulative frequency distribution of BSA (fbsa) in the population so that from the maximal treatable Vsa calcu- lated, the maximal treatable BSA for each PTR and therapy can be computed, and from the cumulative fre- quency distribution of BSA, we can compute the fraction of the total population for which Kt/V urea or C Cr /1.73 m 2 target can be reached with each therapy. The fraction of the total patient population (ftpop) in each PTR group treatable with each therapy is then calculated from the product: A family of kinetic curves was computed for each of the four transport categories (low, low average, high average, and high). The therapies modeled for each transport category were CAPD-8, CAPD-10, PD Plus-12, and PD Plus-15. The PackPD kinetic modeling program, which was previously validated [15, 16], was used with pt50 urea values corresponding to the transport categories to compute weekly Kt/V urea for all four therapies as a function of body water [noted below as volume flow (V) calculated from BSA or volume of distribution of urea (Vsa)]. The pt50 urea value is the dwell time required for the study solute concentration in the peritoneal solution to equal 50% of the plasma concentration. Values for Kt/V urea were computed with the modeling program for each transport category and each therapy with Vsa vary- ing from 20 to 60 liters. The results (Fig. 2) were found to be best expressed as power functions (Table 2) of the general formula: Kt/V urea a (Vsa) b (Eq. 1) where Vsa is expressed in liters. A target weekly prescribed Kpt/V urea of 2.1 was plotted on each therapy curve to demonstrate the maximum patient Vsa possible to attain the prescribed level of therapy. The determination of weekly C Cr for the various PTR groups was not possible using the pooled data because of Assessment of feasibility of achieving adequacy The maximum Vsa that can be successfully treated must be determined from an appropriate solution of equation 1 for Kt/V urea 2.1 for each transport category and each therapy. The power function in equation 1 is converted to exponential form in accordance with 2.1 a * exp(lnvsa * b) (Eq. 2) The solution for equation 2 for Vsa is Vsa exp [ln (2.1/a)/b] (Eq. 3) Thus, with equation 3, we can compute maximum Vsa at which Kt/V urea 2.1 for each transport category and each therapy. When we know the maximum Vsa for which a Kt/V urea of 2.1 can be achieved for each of the four transport groups with each of the four therapies, we can calculate the maximum BSA for which Kt/V urea of 2.1 can be achieved by solving the Hume equations for BSA with known Vsa [17]. These solutions are: BSA (Vsa 8.46)/25.47 for male patients (Eq. 4) BSA (Vsa 7.73)/22.72 for female patients (Eq. 5) ftpop (fptr) (fbsa) The results of these calculations were expressed as a percentage. For example, the percentage of male patients achieving adequacy (weekly Kpt/V of 2.1 or more) with CAPD-10 and who are high-average transporters is determined as follows: Kt/V (Vsa) 1.030

5 Diaz-Buxo et al: Peritoneal dialysis adequacy 2497 Fig. 4. Predicted weekly Kpt/V for a male patient with a BSA 1.73 m 2 and a Vsa 35.6 liters according to peritoneal transport rate and prescription. The values over each bar denote Kpt/V for the specific PTR and prescription. Symbols are: ( ) low; ( ) lowaverage; ( ) high-average; ( ) high. Fig. 5. Predicted maximum body surface area (BSA) for males and females undergoing different prescriptions according to peritoneal transport rate in order to achieve a Kpt/V of 2.1 or more. Values above the bars denote the BSA for the specific PTR, prescription, and gender. Abbreviations are: L, low; LA, low-average; HA, high-average; H, high; M, male; and F, female. Symbols are: ( ) L-M; ( ) L-F; ( ) LA-M; ( ) LA-F; ( ) HA-M; ( ) HA-F; ( ) H-M; ( ) H-F. By using equation 3, BSA ( )/ m 2 Vsa exp [ln (2.1/80.173)/( 1.03)] or the maximum BSA for a male patient with high- Vsa exp [ 36422/ 1.030] liters average transport undergoing CAPD-10 while achieving a Kt/V urea of 2.1. From equation 4, the BSA from the Vsa for male patients The fraction of the total patient population in the can be calculated as follows: high-average PTR (FPTR) is 35.1%, and the cumulative frequency distribution of a BSA of 1.68 m 2 or less (fbsa) BSA (Vsa 8.46)/25.47 is 51.8%. Therefore, applying equation 6, the fraction of

6 2498 Diaz-Buxo et al: Peritoneal dialysis adequacy the total male population with high-average transport DISCUSSION that can achieve a Kt/V urea of 2.1 or less is calculated as: This study is based on kinetic modeling using a validated ftpop (fptr) (fbsa) program and using the actual PTRs and body size ftpop (0.351) (0.518) or 18.2% distribution from a large cross-sectional sample of pa- tients. The modeling was performed assuming no contribution from RRF. The model nonetheless assumes patient RESULTS compliance and the optimal prescription for the specific modality of therapy. The distribution of BSA for patients undergoing HD Patient size, based on BSA, is probably the same for and PD is shown in Figure 1. The mean BSA for HD the PD and HD population. The slightly higher values was 1.74 m 2. For PD, it was 1.80 m 2, and for the combined observed for PD patients could be explained by the population, it was 1.75 m 2. The gender distribution was higher weights of the PD patients resulting from the similar for both groups, with 51% males among HD and dialysate carried by some patients. Our results would 49% among PD patients. not be significantly affected by using patient size based Peritoneal transport groups are defined in Table 1, on the PD populations, the HD population, or the comand the peritoneal transport curves are defined in Table bined results (used in this study) because the difference 2. Based on the D/P Cr at four hours, 14.1, 34.3, 35.1, and in BSA was very small at 0.06 m 2, and the BSA distribu- 16.4% of the patients fell into the low, low-average, tion histograms categorized patients by 0.10 m 2. high-average, and high categories, respectively. Figure 2 A possible source of error is the derivation of Vsa summarizes the weekly Ktp/V modeled from the from the Hume equation rather than from the Watson PackPD data pool by patient size (Vsa) and dialysis model [18], which considers age and sex, in addition to prescription for each PTR group. The mathematical weight and height. These methods of calculating volume definition of the various peritoneal transport curves is of distribution of urea generally agree well; however, provided in Table 2. A consistent upward shift of the they may differ by as much as 10%. When these methods curves is observed for every increase in PTR. Figure 3 differ, it is often not clear which is more accurate. plots C Cr obtained with different prescriptions in patients The PTR distribution is, by definition, quasi-normal, with different PTRs. and the categories are very similar to those reported by Figure 4 shows the Kpt/V predicted for a male patient Twardowski et al in a much smaller sample of patients, with a BSA 1.73 m 2 and a Vsa 35.6 l according to particularly for D/P Cr [12]. PD modality and for different PTRs. Figure 5 illustrates The model predicts that the great majority of patients the predicted maximum BSA for males and females un- (more than 90%) can achieve adequacy levels for dergoing PD with different modalities and with various Kt/V urea in the absence of RRF and that most patients PTRs in order to achieve a weekly Kpt/V of 2.1 or more. (more than 70%) can also satisfy the C Cr criterion with The BSA is consistently higher for females for the same the available therapeutic armamentarium. The large PTR group and prescription because of their lower Vsa variation noted in the curves describing C Cr among high relative to males with the same BSA. transporters undergoing PD Plus may reflect the limited The percentage of patients capable of achieving a number of observations. It is important to stress that weekly Kpt/V of 2.1 or more with the various prescriponly a small proportion of the ESRD population can be tions and PTRs are provided for males (Fig. 6), females (Fig. 7), and all patients (Fig. 8). Standard CAPD-8 only adequately dialyzed with standard CAPD in the absence satisfied the adequacy criterion in 24.8% of cases, whereas of RRF, regardless of the criterion of adequacy used. PD Plus-15 achieved the goal in 93.2% for the combined Although CAPD is an important and very commonly group of all patients. used form of therapy for the initiation of dialysis, fre- The percentage of patients able to achieve a weekly quent monitoring of RRF and adjustments in dialysis C Cr of 60 liter/1.73 m 2 or more with the various prescrip- dose are needed to maintain adequacy. tions and PTRs is shown in Figure 9. The proportion of Several important adequacy issues currently remain patients capable of achieving adequacy fell by 20% using unsolved. The current standards of adequacy are based the C Cr of 60 liter/1.73 m 2 or more criterion of adequacy. on the assumption that peritoneal and renal clearance Figure 10 illustrates the predicted maximum BSA for are equivalent. If prospective, controlled clinical studies patients undergoing PD with different modalities and were to show that the renal contribution is significantly with various PTRs in order to achieve a weekly C Cr of better than peritoneal clearance, it is reasonable to as- 60 liter/1.73 m 2 or more. A strong correlation between sume that the goals of adequacy should be higher for dialysis solution volume and the proportion of patients anuric patients. The importance of urea or creatinine as achieving adequacy with either criterion was observed markers of clinical outcome in ESRD is also unclear. (Fig. 11). The correlation between clinical outcome and PD dose

7 Diaz-Buxo et al: Peritoneal dialysis adequacy 2499 Fig. 6. Predicted proportion of male patients capable of achieving a weekly Kpt/V of 2.1 or more according to modality of therapy and peritoneal transport rates. Values above the bars denote the percentage of patients achieving Kpt/V for that specific PTR group and prescription. The total percentage refers to the total proportion of male patients predicted to attain a Kpt/V 2.1 with that particular prescription. Symbols are: ( ) low; ( ) lowaverage; ( ) high-average; ( ) high. Fig. 7. Predicted proportion of female patients capable of achieving a weekly Kpt/V of 2.1 or more according to modality of therapy and peritoneal transport rates. Symbols are: ( ) low; ( ) low-average; ( ) high-average; ( ) high. Fig. 8. Predicted proportion of all patients capable of achieving a weekly Kpt/V of 2.1 or more according to modality of therapy and peritoneal transport rates. Symbols are: ( ) low; ( ) low-average; ( ) high-average; ( ) high.

8 2500 Diaz-Buxo et al: Peritoneal dialysis adequacy Fig. 9. Predicted proportion of patients capable of achieving a weekly C Cr of 60 liter/1.73 m 2 or more according to modality of therapy and peritoneal transport rates. Symbols are: ( ) low; ( ) low-average; ( ) high-average; ( ) high. Fig. 10. Predicted maximum body surface area (BSA) for patients undergoing different modalities according to peritoneal transport rate in order to achieve a C Cr of 60 liter/ 1.73 m 2 or more. Symbols are: ( ) low; ( ) low-average; ( ) high-average; ( ) high. Fig. 11. Relationship between total daily volume of dialysis solution used and degree of adequacy achieved according to the two adequacy criteria used: ( ) weekly Kpt/V urea 2.1 and ( ) C Cr 60 liter/1.73 m 2. Solid lines without symbols represent linear regressions. suggests that urea indices have more favorable results than creatinine [4, 5]. Trying to satisfy the creatinine clearance goals, however, is recommended because that effort will guarantee a higher total dialysis dose. It is much easier to satisfy the urea clearance adequacy goals than the creatinine, except in very high transporters. Finally, several studies have strongly suggested that high peritoneal transport is associated with inferior prognosis [19 24]. If faster transport is indeed a marker of poor clinical outcome, justifying transfer of those patients to HD, the proportion of patients able to dialyze adequately with PD would diminish significantly. Thus, it is important to further characterize the potential etiology of rapid peritoneal transport and to develop effective means to prevent and treat protein malnutrition, which is common among high transporters. The higher suggested standards of PD therapy and the aforementioned unresolved adequacy issues favor

9 Diaz-Buxo et al: Peritoneal dialysis adequacy 2501 the use of APD. As RRF is lost, most patients require clinical practices for maximizing peritoneal dialysis clearances. Perit Dial Int 16(Suppl 5):S448 S456, 1996 higher doses of therapy not achievable with CAPD. The 9. Burkart JM, Schreiber M, Korbet SM, Churchill DN, Hamburger current worldwide trend confirms an increased utilizaapproach RJ, Moran J, Soderbloom R, Nolph KD: Solute clearance to adequacy of peritoneal dialysis. Perit Dial Int 16:457 tion of APD modalities [25]. The clinical evidence accu- 470, 1996 mulated during the last decade has stimulated the search 10. Rocco MV: Body surface area limitations in achieving adequate for more effective modalities of therapy characterized therapy in peritoneal dialysis patients. Perit Dial Int 16: , 1996 by higher dialysate flow rate (DFR), automated delivery, 11. du Bois D, du Bois EF: A formula to estimate the approximate and hopefully shorter procedural time. surface area if height and weight be known. Arch Intern Med 17: , 1916 ACKNOWLEDGMENTS 12. Twardowski ZJ, Nolph KD, Khanna R, Prowant BF, Ryan LP, Moore HL, Nielsen MP: Peritoneal equilibration test. Perit Dial Bull 7: , 1987 The preliminary results from this work were presented at the 18th 13. Cook JGH: Factors influencing the assay of creatinine. Ann Clin Annual Conference on Peritoneal Dialysis, Nashville, Tennessee, on Biochem 12: , 1975 February 25, 1998, and at the 3rd European Peritoneal Dialysis Meet- 14. Diaz-Buxo JA: Enhancement of peritoneal dialysis: The PD Plus ing, Edinburgh, Scotland, on April 5, concept. Am J Kidney Dis 27:92 98, Gotch FA, Lipps BJ, Pack PD: A urea kinetic modeling computer Reprint requests to Jose A. Diaz-Buxo, M.D., Fresenius Medical program for peritoneal dialysis. Perit Dial Int 17(Suppl 2):S126 Care NA, Peritoneal Dialysis Services, 1051 East Morehead Street, Suite S130, , Charlotte, North Carolina 28204, USA. 16. Gotch FA, Lipps BJ, Keen ML, Panlilio F: Coinvestigators for pdserve@bellsouth.net the Fresenius Randomized Dialysis Prescriptions and Clinical Outcome Study (RDP/CO): Computerized urea kinetic modeling to REFERENCES prescribe and monitor delivered Kt/V (pkt/v, dkt/v) in peritoneal dialysis, in Advances in Peritoneal Dialysis, edited by Khanna R, Toronto, Multimed Inc., 1996, pp Gotch FA, Keen ML: Kinetic modeling in peritoneal dialysis, in 17. Hume R, Weyers E: The relationship between total body water Clinical Dialysis, edited by Nissenson AR, Fine RN, Gentile DE, and surface area in normal and obese subjects. J Clin Pathol 24:234 East Norwalk, Appleton & Lange, 1995, pp , CANADA-USA (CANUSA) Peritoneal Dialysis Study Group: 18. Watson PE, Watson ID, Batt RD: Total body water volumes Adequacy of dialysis and nutrition in continuous peritoneal dial- for adult males and females estimated from simple anthropometric ysis: Association with clinical outcomes. J Am Soc Nephrol 7:198 measurements. Am J Clin Nutr 33:27 39, , Nolph KD, Moore HL, Prowant B, Twardowski ZJ, Khanna 3. NKF DOQI Clinical Practice Guidelines for Peritoneal Dialysis R, Gamboa S, Keshaviah P: Continuous ambulatory peritoneal Adequacy. Am J Kidney Dis 30(Suppl 3):S67 S136, 1997 dialysis with a high flux membrane. ASAIO J 39: , Selgas R, Bajo MA, Fernandez-Reyes MJ, Bosque E, Lopez- 20. Churchill DN, Thorpe KE, Nolph KD, Keshaviah PR, Oreo- Revuelta K, Jimenez C, Borrego F, de Alvaro F: An analysis poulos DG, Pagé D: Increased peritoneal membrane transport of adequacy of dialysis in a selected population on CAPD for over is associated with decreased patient and technique survival for 3 years: The influence of urea and creatinine kinetics. Nephrol continuous peritoneal dialysis patients. J Am Soc Nephrol 9:1285 Dial Transplant 8: , , Mehrotra R, Saran R, Nolph KD, Moore HL, Khanna R: Evi- 21. Blake PG: What is the problem with high transporters? Perit Dial dence that urea is a better surrogate marker of uremic toxicity Int 17: , 1997 than creatinine. ASAIO J 43:M858 M861, Fried L: Higher membrane permeability predicts poorer patient 6. Gotch FA: The CANUSA study. Perit Dial Int 17(Suppl 2):S111 survival. Perit Dial Int 17: , 1997 S114, Davies SJ, Phillips L, Russell GI: Peritoneal solute transport predicts survival on CAPD independently of residual renal func- 7. Gotch FA, Gentile DE, Keen ML, Amerling R, Folkert VW, tion. Nephrol Dial Transplant 13: , 1998 Kliger AS, Shapiro WB: The incident patient cohort study design 24. Wang T, Heimburger O, Waniewski J, Bergstrom J, Lindholm with uncontrolled dose: Substantial over-estimation of mortality B: Increased peritoneal permeability is associated with decreased as a function of peritoneal dialysis dose? ASAIO J 42:M514 M517, fluid and small-solute removal and higher mortality in CAPD pa tients. Nephrol Dial Transplant 13: , Blake P, Burkart JM, Churchill DN, Daugirdas JT, Depner 25. U.S. Renal Data System: USRDS 1997 Annual Data Report. TA, Hamburger RJ, Hull AR, Korbet SM, Moran J, Nolph KD, Bethesda, National Institutes of Health, National Institute of Dia- Oreopoulos DG, Schreiber M, Soderbloom R: Recommended betes and Digestive and Kidney Diseases, 1997